NLR is an established prognostic predictor in various human cancers (9). For breast cancer, a systematic review conducted by Corbeau et al. (10) revealed that the majority of studies showed negative prognostic correlation between all survival outcomes (DFS, BCSS and OS) and a high NLR (11–13), including in cohorts of patients receiving NAC (14, 15). TIL is a predictor of response to NAC (16), and a robust prognostic factor for both treatment-naive and post-neoadjuvant chemotherapy breast cancers (17–19), and is recognized as an important histologic feature for breast cancer classification in the latest World Health Organization classification of tumours of the breast (20). There is possible correlation between circulating (NLR) and local (TLR) immune microenvironment, supported by the strong association between high TIL and high circulating lymphocyte levels (i.e., low NLR) with favorable outcomes.
Findings from the current study reinforces the positive correlation between high TIL and low NLR in the neoadjuvant setting (Table 3). The cutoff for adopted high and low NLR was 2.6 in this study. In other cohorts of patients receiving NAC, cutoffs for NLR ranged from 1.7 to 3.3 (21, 22) but the majority were between 2.0 to 2.9 (14, 15, 23). Previous studies have reported inferior clinical outcomes for patients with high pre-chemotherapy (baseline) NLR (21, 22). In the current cohort, OS, BCSS and DFS for pre-chemotherapy NLR-high patients were all shorter but did not reach statistical significance (Table 1). Studies analyzing post-chemotherapy and late post-treatment NLR, have reported high-NLR after treatment to be indicative of shorter survival (24, 25). Kim et al. demonstrated association between an increase of NLR at one year post-surgery with poor prognosis (26). In line with the literature, all outcome metrics (OS, BCSS and DFS) were significantly worse in the NLR-high group for this cohort. An increase of NLR of more than 0.5 from pre-chemotherapy to late post-surgery was found to indicate worse DFS and BCSS, whereas for post-chemotherapy NLR-high cases, such an increase was associated with improved OS and BCSS (supplementary Fig. 1).
Evidence regarding the relationship between NLR and TILs is less established. Cohorts analyzing NLR and TIL in colorectal and ovarian cancer patients did not find correlations between NLR and TILs (4, 5, 8), whereas a negative correlation was demonstrated for biliary tract cancers (27). Regardless, all of those studies showed high-NLR and low-TIL to be associated with worse prognosis (4, 5, 8, 27). As for breast cancer, most studies on the relationship of NLR and TILs only included pre-treatment NLR (6, 7, 28, 29). Although high-NLR and low-TIL continue to demonstrate favorable prognosis (6, 7, 28, 29), correlation between NLR and TIL levels were not found (6, 7). One study by Lee et al. (30) which included pre-treatment and post-chemotherapy NLR reported that a decrease in NLR and same/increased TILs to be associated with prolonged DFS, but correlation analysis between NLR with TILs was not performed. An ongoing clinical trial aiming to evaluate NLR vs. TILs is currently underway (31). In the current study, TILs on H&E slides, similar to previous studies, was not associated with pre-treatment NLR, but demonstrated negative correlation to late post-surgery NLR. The addition of CD8 IHC showed that both stromal and tumor CD8 TIL levels were lower in pre-chemotherapy NLR-high cases (Table 2). These findings support a correlation between circulating and local lymphocyte levels in breast cancers.
High pre-chemotherapy NLR was associated with necrosis on biopsy and excision, whereas high post-chemotherapy NLR was associated with necrosis on excision (Table 1). TAMs were also more often seen in cases with high pre-chemotherapy NLR (Table 1). The positive association between NLR, tumor necrosis and TAMs has been reported in hepatopancreatic tumors (32, 33). But for breast cancers, literature search only yielded a smaller cohort which failed to demonstrate associations between NLR and tumor necrosis in breast cancer (34). It has been suggested that cancer-associated inflammatory cytokines, for example tumor necrosis factor alpha and colony stimulating factor contribute to the increase of neutrophil count (1, 35). High NLR reflects increased inflammatory activity attributable to cancer burden, proliferation and/or progression, which in turn is prognostically significant (1, 2). This postulation reconciles with the finding that a late increase in NLR is associated with poor outcomes (36). Interestingly, a further increase in NLR for cases with high NLR post-chemotherapy is indicative of longer survival. It is possible that a high post-chemotherapy NLR suggests residual disease, and further increases signifies an active and favorable immune response.
Multivariate analysis revealed that late-post surgery NLR was an independent negative predictive factor for OS, BCSS and DFS. Biomarker status and nodal staging were also consistently correlated with OS, BCSS and DFS. TIL showed independent correlation only for OS and BCSS but not DFS (Table 3). TIL, nodal status, hormone receptor markers and HER2 expression are recognized prognostic factors for breast cancer (37). With reference to available literature, biomarker status, tumor staging and grading, but not TILs, have been reported as independent prognostic factors to NLR (10, 14). Although NLR and TILs are correlated, there appears to be a certain degree of independency and separate mechanisms of interaction between the local (NLR) and circulating (TIL) immune microenvironment with breast cancer and its prognosis.
Most studies on populations receiving NAC compared pre-chemotherapy NLR only, and although negative correlation with survival were seen, multivariate analysis did not consistently demonstrate independent association (14, 15, 21, 38, 39). Two studies involving multiple time points reported significant association between 6-month and 24-to-36-month post-chemotherapy NLR with survival, but not pre-treatment NLR (24, 25). The current cohort echoes these two studies and suggests that post-chemotherapy NLR may be a stronger predictor of prognosis than pre-chemotherapy NLR. By analyzing multiple time points, it appears that a later NLR is more prognostically relevant. It may be possible that late high NLR levels signifies residual disease activity and recurrence (40).